DeepGroveYC S25 landing page report.
← Back to YC S25

Your landing page’s weakest score is Trust.

Fix Trust

Score over time

807060504030Liquid AI 71 on May 28 (Wayback)Liquid AI 72 on Jun 15 (Wayback)Liquid AI 67 on Aug 5 (Wayback)Liquid AI 70 on Aug 13 (Wayback)Liquid AI 68 on Aug 28 (Wayback)Liquid AI 69 on Sep 17 (Wayback)Liquid AI 73 on Sep 22 (Live)Nexa AI 49 on Apr 4 (Wayback)Nexa AI 50 on Apr 11 (Wayback)Nexa AI 57 on Sep 22 (Live)Mistral AI 70 on May 16 (Wayback)Mistral AI 60 on Jun 14 (Wayback)Mistral AI 63 on Jul 25 (Wayback)Mistral AI 62 on Aug 7 (Wayback)Mistral AI 65 on Aug 23 (Wayback)Mistral AI 53 on Sep 20 (Live)Cartesia 64 on Jun 3 (Wayback)Cartesia 64 on Jun 12 (Wayback)Cartesia 68 on Jul 27 (Wayback)Cartesia 66 on Aug 21 (Wayback)Cartesia 66 on Sep 4 (Wayback)Cartesia 69 on Sep 20 (Live)Kyutai 53 on May 28 (Wayback)Kyutai 55 on Jun 16 (Wayback)Kyutai 52 on Jun 29 (Wayback)Kyutai 52 on Jul 11 (Wayback)Kyutai 52 on Jul 21 (Wayback)Kyutai 54 on Aug 25 (Wayback)Kyutai 57 on Sep 22 (Live)DeepGrove 43 on Aug 11 (Wayback)DeepGrove 46 on Sep 18 (Wayback)DeepGrove 48 on Sep 22 (Live)Aug 11Sep 18Sep 22
Past reportsShow past reportsHide past reports
  1. Sep 22, 202648LiveOverall +2 · CTA +7 · Information hierarchy +6Viewing
  2. Sep 18, 202646Overall +3 · Trust +16 · Mobile web −6View →
  3. Aug 11, 202643First tracked report — strongest Mobile web 58, weakest Trust 22View →

Trust

34Poor

Proof that the product is real, competent, and safe to try.

DeepGrove
Liquid AI
Kyutai
Cartesia
Mistral AI
Trust read
DeepGrove"GitHub" link in the top navigation
Liquid AI"56 LFMS SHIPPED / 47.7M DOWNLOADS / 3500+ VARIANTS" stat block above the primary CTA
Nexa AIimplicit context Nexa AI may benefit from pre-existing familiarity among its target audience.
Mistral AIFeatured News card: "Mistral raises €3B to make sovereign, open-weight AI the…"
CartesiaInteractive "Talk to Skylar" voice demo carousel
KyutaiThree named live demos — "MIRA," "Moshi," and "Unmute" — with concrete descriptions like "Real-time voice AI with natura

Value props

48Poor

Title, subtitle, hero — do they sell the product's true value?

DeepGrove’s headline vagueFrontier Intelligence On Any Device

Working

  • Hints at on-device differentiator
  • Short and memorable
  • Consistent across desktop and mobile

Could improve

  • No subhead or supporting proof
  • "Frontier Intelligence" is generic
  • Art decorates, demonstrates nothing

Mobile web

48Poor

The phone experience as a first-class citizen.

Mobile web annotations on the mobile capture12
  1. Maple Launch link — Primary action is a small gray text link near the page bottom, not a thumb-friendly button in the first viewport.
  2. Bottom-anchored headline — "Frontier Intelligence On Any Device" sits roughly two viewports down, so the first screen makes no value statement.

Information hierarchy

54Fair

Does the design help or inhibit what the page is trying to say?

Information hierarchy annotations on the desktop capture1234
  1. Headline pinned at fold — "Frontier Intelligence On Any Device" sits at the very bottom edge, forcing visitors to process an unlabeled line field before reading the value prop.
  2. Chat button — The primary "Chat" CTA is a small pill in the top-right corner, far from the headline and outside the natural scan path of the sparse fold.
  3. Maple Launch link — A second same-weight action beside the headline competes with "Chat" and points attention in a different direction with no context.
  4. Nav link row — Four exit links (Research, Careers, Maple-Preview, GitHub) styled at near-equal weight to "Chat" leak attention before the value prop is even read.

CTA

55Fair

Copy, size, verb, cost-feel — does the primary CTA earn the click?

DeepGrove
Cartesia
Liquid AI
Nexa AI
Mistral AI

Problems

  • no action verb — reads as a label
  • tiny nav pill, no hero CTA
  • hidden behind hamburger on mobile
  • hero offers nothing to click

Competitors

inferred from product, audience, and value prop